Printed Thai character recognition using the hierarchical cross-correlation ARTMAP

Abstract

Traditionally, Thai characters are composed of circle, zigzag line, curve, and head. However, many new Thai fonts, which are now gaining in popularity, do not follow the traditional writing rule; the head has been omitted from the characters. Without the head, it is very difficult to segregate the characters. Even the best commercial Thai OCR software has difficulty in recognizing this kind of character. Therefore, the hierarchical cross-correlation ARTMAP is proposed in this paper to recognize the no-head Thai characters.

Mixed Thai-English Character Classification Based on Histogram of Oriented Gradient Feature

“…First, Thammano and Duangprasuk [6] suggested a way to extract Thai-character features by separating a character image into 3x3 cell then finding a starting point and tracing the direction of the connecting points within the cell. However, the feature extracted hardly uses the characteristic of Thai characters.…”

“…Add to that, many documents written in Thai can sometimes contain A typical character recognition system, however coupling might it be, can be separated into two tasks: a feature extraction and a classification. For the classification, there are published papers utilizing Hidden Markov Model [1], [2], Artificial Neuron Network [3], [4], [5], [6], [7], fuzzy rough set [8], Karhunen-Loeve expansion [9], as well as a hybrid approach [10]. Evidently, there are vast varieties of classification tools which can be used for an OCR system.…”

Printed Thai Character Recognition using Shape Classification in Video Sequence along a Line

“…We will use these appearances to classify characters into four main groups as the following first [4,5].…”

“…We test our recognition system with text image sequences in 4 fronts which are; Angsana New, Cordia New, Tahoma and TH Sarabun New. Our recognition system consist of 3 main processes which are: (1) The preparing of image sequences before character recognition process [1]; (2) Recognition process of each character; [2][3][4][5] and (3) The grouping of characters in each line and display text recognition [1].…”

Character classification framework based on support vector machine and k-nearest neighbour schemes

“…Classification frameworks were reported based on artificial neural networks [8][9][10] , fuzzy rough sets 11 , as well as a hybrid approach based on fuzzy membership function and neural networks 12 . However, these studies train their classification tool(s) by grouping all fonts together which can cause a reduction in the separability of characters.…”